Skip to content

[PyTorch] Relax dimension constraints for using fused grouped MLP #2856

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
ksivaman wants to merge 5 commits into
base: main
Choose a base branch
from
Open

[PyTorch] Relax dimension constraints for using fused grouped MLP #2856

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
638c351
Reduce fused path dim constraint
ksivaman Apr 2, 2026
7303c60
Merge branch 'main' into relax_dim_constraint_for_fused_grouped_mlp
ksivaman Apr 9, 2026
8cdc033
Fix randomization in tests
ksivaman Apr 9, 2026
51e7e6f
Merge branch 'main' into relax_dim_constraint_for_fused_grouped_mlp
ksivaman Apr 9, 2026
f3d6950
reset rng as before, assert input dim
ksivaman Apr 10, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion tests/pytorch/test_fusible_ops.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3249,12 +3249,13 @@ def to_cpu(tensor: Optional[torch.Tensor]) -> Optional[torch.Tensor]:
@pytest.mark.parametrize("accumulate_into_main_grad", (False, True))
@pytest.mark.parametrize("glu_interleave_size", (None, 32))
@pytest.mark.parametrize("delay_wgrad_compute", (False, True))
@pytest.mark.parametrize("hidden_size", (64, 256))
def test_grouped_mlp(
self,
*,
group_size: int = 4,
bias: bool,
hidden_size: int = 256,
hidden_size: int,
dtype: torch.dtype,
quantization: Optional[str],
single_grouped_weight: bool,
Expand Down
12 changes: 6 additions & 6 deletions transformer_engine/pytorch/ops/_common.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,12 +76,12 @@ def get_fp8_meta_from_fp8_tensor(tensor: Float8Tensor) -> tuple[FP8TensorMeta, i
def validate_grouped_mlp_dims(fc1, swiglu, fc2) -> None:
"""Validate FC1/SwiGLU/FC2 dimensions and interleave size for fused grouped MLP."""

if fc1.in_features % 256 != 0 or fc1.out_features % 256 != 0:
if fc1.in_features % 64 != 0 or fc1.out_features % 64 != 0:
raise ValueError(
f"Unsupported dims for FC1 (num_groups={fc1.num_groups}, "
f"in_features={fc1.in_features}, out_features={fc1.out_features})."
)
if fc2.in_features % 256 != 0 or fc2.out_features % 256 != 0:
if fc2.in_features % 64 != 0 or fc2.out_features % 64 != 0:
raise ValueError(
f"Unsupported dims for FC2 (num_groups={fc2.num_groups}, "
f"in_features={fc2.in_features}, out_features={fc2.out_features})."
Expand Down Expand Up @@ -153,10 +153,10 @@ def fuse_grouped_mlp_ops(
elif window[0].num_groups != window[2].num_groups:
matches_pattern = False
elif (
window[0].in_features % 256 != 0
or window[0].out_features % 256 != 0
or window[2].in_features % 256 != 0
or window[2].out_features % 256 != 0
window[0].in_features % 64 != 0
or window[0].out_features % 64 != 0
or window[2].in_features % 64 != 0
or window[2].out_features % 64 != 0
):
matches_pattern = False
elif window[1].glu_interleave_size != 32:
Expand Down
12 changes: 6 additions & 6 deletions transformer_engine/pytorch/ops/fused/backward_grouped_mlp.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -401,8 +401,8 @@ def fuser_backward(
fc2_dy_scales = fc2_dy_scales.view(dtype=torch.float8_e8m0fnu)
fc2_dy_scales = fc2_dy_scales.view(
1,
out_shape[0] // 128,
out_shape[1] // 128,
(out_shape[0] + 127) // 128,
(out_shape[1] + 127) // 128,
MXFP8_BLOCK_SCALING_SIZE,
4,
4,
Expand Down Expand Up @@ -451,8 +451,8 @@ def fuser_backward(
fc2_w_scales = fc2_weight_for_gemm.columnwise_scale_inv.view(dtype=torch.float8_e8m0fnu)
fc2_w_scales = fc2_w_scales.view(
num_groups,
fc2_weight_shape[1] // 128,
fc2_weight_shape[0] // 128,
(fc2_weight_shape[1] + 127) // 128,
(fc2_weight_shape[0] + 127) // 128,
MXFP8_BLOCK_SCALING_SIZE,
4,
4,
Expand Down Expand Up @@ -582,8 +582,8 @@ def fuser_backward(
)
fc1_w_scales = fc1_w_scales.view(
num_groups,
fc1_weight_shape[1] // 128,
fc1_weight_shape[0] // 128,
(fc1_weight_shape[1] + 127) // 128,
(fc1_weight_shape[0] + 127) // 128,
MXFP8_BLOCK_SCALING_SIZE,
4,
4,
Expand Down
12 changes: 6 additions & 6 deletions transformer_engine/pytorch/ops/fused/forward_grouped_mlp.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -309,8 +309,8 @@ def fuser_forward(
fc1_x_scales = fc1_x_scales.view(dtype=torch.float8_e8m0fnu)
fc1_x_scales = fc1_x_scales.view(
1,
in_shape[0] // 128,
in_shape[1] // 128,
(in_shape[0] + 127) // 128,
(in_shape[1] + 127) // 128,
MXFP8_BLOCK_SCALING_SIZE,
4,
4,
Expand Down Expand Up @@ -358,8 +358,8 @@ def fuser_forward(
fc1_w_scales = fc1_weight_for_gemm.scale_inv.view(dtype=torch.float8_e8m0fnu)
fc1_w_scales = fc1_w_scales.view(
num_groups,
fc1_weight_shape[0] // 128,
fc1_weight_shape[1] // 128,
(fc1_weight_shape[0] + 127) // 128,
(fc1_weight_shape[1] + 127) // 128,
MXFP8_BLOCK_SCALING_SIZE,
4,
4,
Expand Down Expand Up @@ -455,8 +455,8 @@ def fuser_forward(
fc2_w_scales = fc2_weight_for_gemm.scale_inv.view(dtype=torch.float8_e8m0fnu)
fc2_w_scales = fc2_w_scales.view(
num_groups,
fc2_weight_shape[0] // 128,
fc2_weight_shape[1] // 128,
(fc2_weight_shape[0] + 127) // 128,
(fc2_weight_shape[1] + 127) // 128,
MXFP8_BLOCK_SCALING_SIZE,
4,
4,
Expand Down
Loading